[Gnu-arch-users] Re: OT: Lisp

[Tom Lord]

    > From: Miles Bader <address@hidden>

    > One of lisp's strengths is that the two (`language tinkering'
    > and `being a productive tool') aren't at all incompatible, and
    > can take place simultaneously.


!!!!! yes.

This is something that has happened during the "history of computing":
we've become way more compartmentalized than we used to be.   We have
this mythology and ethos and even legal structure of "os level"
vs. "language level" vs. "application level" and if you're hacking at
one of those you can't hack at the others.

It didn't used to be that way.  Back when there was _way_ much less
software -- you'd just hack at whatever level you needed today and
whatever different level you needed to tomorrow.

It's easy to brush off this loss of freedom as "managing complexity"
but I'm not so sure I really buy it.   It's more like "how we have
been packaged or expect to be packaged."


    > Some reasons I can see for this:

    >   (1) Any `language tinkering' takes place within the existing low-level
    >       syntax, and so doesn't disrupt unrelated code (unlike changing the
    >       low-level syntax)

    >   (2) There are well-established conventions for making new syntax

    >   (3) It's really, really, really, simple

    > The result is that if your life would be made easier by some special
    > syntax, you can just get down and do it, without spending a lot of time
    > dicking around with the details, and you're a lot more likely to do it
    > right the first time.

Don't leave out the earlier days when there were things like:

        (4) you can open the kernel source in an emacs window and
            hack it live

        (5) in a desperate pinch, you can always hack the microcode


(c.f. the thread about bugs in bash.)

-t